Obtaining CTSA pilot funding

Dr Crew described the pilot funding she received from the Irving Institute – funding specifically targeted to promoting collaborative and multidisciplinary research – as being the most instrumental for the development of RealRisks and BNAV. Specifically, this funding was critical for enabling Dr Crew to develop new, interdisciplinary partnerships – including her partnership with Dr Kukafka – and in pushing the bounds of her research to include new methods (e.g., community-based participatory research) and new populations (e.g., Ashkenazi Jewish women). These initial pilot awards facilitated data collection to show the initial promise of the intervention/research program and apply for subsequent grant funding. The result is an interdisciplinary team, consisting of oncologists, epidemiologists, software developers, data scientists, and public health experts.

Interdisciplinary collaborations

Interview participants pointed to the centrality of building strong, interdisciplinary relationships to catalyze the development of clinical interventions and research programs. Although creating interdisciplinary teams requires patience, there was a recognition that the development of RealRisks and BNAV would not have been possible without robust collaboration between scientists from multiple disciplines working toward a common goal. Drs. Crew and Kukafka both reflected their sense that solving complex health problems requires a team approach, with Dr Kukafka reflecting, “health conditions don’t have a single etiology and neither do solutions.” They both felt their work was enriched by cross-fertilizing their thinking with people from other disciplines but noted that engaging in this type of work is partly a matter of personal interest (e.g.., according to Dr Kukafka, “I think there are some people who are inherently interested in [interdisciplinary work] and some people who are not”) and partly a result of infrastructures that support such interdisciplinary collaborations. In this respect, the development of RealRisks and BNAV was facilitated not only by funding to support interdisciplinary collaboration, but also by forums created by the Irving Institute and Columbia University Irving Medical Center in which researchers could come together from different areas of specialty, engage in opportunities for dialog, share ideas, and identify potential partnerships [Reference Humensky, Abedin and Muhammad24]. For example, Drs. Crew and Kukafka participated in and presented at the Columbia University Herbert Irving Comprehensive Cancer Center’s Interdisciplinary Research Series seminar in January 2021. This focus on interdisciplinarity at the institutional level aligns with vast literature supporting the importance of interdisciplinary work in promoting creativity and novel scientific discoveries [Reference Brint25–Reference Brint, Turk-Bicakci, Proctor and Murphy27].

Building relationships with outside researchers and groups was also critical. Dr Crew became involved in the SWOG Cancer Research Network in 2005 as a junior faculty member through one of her mentors. Over time, she continued to develop her connections to SWOG and take on leadership positions, becoming an executive officer for SWOG’s prevention committee in 2015. This proved vital for future research and SWOG is now the primary collaborator on the current multisite trial.

Support from patient advocacy groups

Patient advocates are a standard part of review committees, and reaching out to patient advocacy groups in designing and implementing RealRisks was critical for its development. Drs. Crew and Kukafka conducted usability studies of their tools with actual end users prior to launch to ensure that the interventions were feasible to use and acceptable to patients of different backgrounds [Reference Kukafka, Fang, Vanegas, Silverman and Crew16]. Dr Crew noted that even as a clinician who has direct care responsibility for patients, talking to trained patient advocates provided a new avenue for understanding issues that patients may not want to share with their clinicians as well as for anticipating how the intervention would be received at a local and national level.

Limited mid-career support

Keeping research funded is an ongoing challenge for scientists but can be particularly difficult for mid-career faculty [Reference Keating, Jasper and Musuuza28]. While Dr Crew benefitted from several career development awards that informed her evolution as a cancer researcher, her work to develop RealRisks and BNAV came alongside her transition from mentored research to independent research. This meant she had less protected career development time and relatively fewer funding opportunities for the development and testing of the decision tool. In turn, Dr Crew became involved in multiple collaborations that could provide salary support – broadening the network for potential future collaborations but also reducing the amount of time available to focus on RealRisks and BNAV as it relates to chemoprevention.

Ongoing costs for technology

A separate issue – also related to funding – is that a web-based decision support tool requires ongoing financial investment, including funding to maintain secure servers and provide security updates. Drs. Crew and Kukafka have been able to maintain continuous support for RealRisks and BNAV by having multiple funders and expanding the reach of the tool beyond its initial focus on chemoprevention. However, given the precariousness of relying on federal funding – and the risk that a lapse in funding would mean shutting down the decision tools – they have started to explore other options for the future including potential paths to commercialization.

Establishing interdisciplinary, team science efforts

The development of RealRisks and BNAV required the collaboration across multiple disciplines. As one interviewee reflected, “If you’re trying to do something new or novel, you need people with a lot of different expertise, disease-based expertise, population science expertise, expertise in implementation science, and expertise in biomedical informatics. You need the whole spectrum, especially when you’re talking about developing new interventions and testing new interventions.” While such team science efforts are recognized as critical, they also present logistical and conceptual challenges. Not only are team members involved in multiple projects and being pulled in different directions, but their diverse backgrounds and areas of expertise can also make it hard to find a common language. Further, current promotion criteria do not always reward team science efforts, which can be a barrier to finding collaborators. Finding other scientists who are, as Kukafka described in her interview, “working at the edge of their discipline” requires patience and willingness. However, Drs. Crew and Kukafka have found that as their work has gained more visibility, other researchers who are interested in interdisciplinary work have sought them out for collaboration. By bringing in different domain experts to fill gaps in knowledge and respecting the expertise of each team member, they continue to expand the network of potential research partners with whom they can collaborate.

Integrating decision tools into clinical workflow

A final challenge in the translation of RealRisks and BNAV has been the integration of the tools at the practice level. As Crew and Kukafka’s team has documented in their qualitative work during tool development, providers face competing demands on their time during any clinical encounter and must establish priorities based on their patients’ needs. Further, primary care providers lack knowledge about chemoprevention and seldom engage in risk communication with their patients (particularly in time-constrained settings) [Reference Jones, Silverman and Guzman29]. Dr Kukafka explained how it can be challenging to “introduce prevention into a health care system that is more episodic and designed to manage acute conditions… it’s not a small challenge – how does it fit into the clinical work flow?” Responding to this challenge and initial results about uptake of BNAV among providers, the research team has since integrated BNAV into the EHR and is studying this integration in the RCT underway.

Current status and future directions

With the current multisite RCT of RealRisks and BNAV almost completed, Dr Crew and Kukafka’s team will soon have more information about the effectiveness of the decision tools in increasing chemoprevention-informed choice as well as their impact on secondary outcomes including chemoprevention knowledge, perceived breast cancer risk/worry, decision conflict, shared decision-making, and chemoprevention uptake and adherence. Results have the potential to inform widespread dissemination of the tool to improve informed, shared decision-making about breast cancer chemoprevention in line with current guidelines.

In the meantime, Drs. Crew and Kukafka have continued their work to add, adapt, and test new modules and enhancements for RealRisks and BNAV. Recent enhancements focus on patient activation. With the current integration of the tool into the EHR, RealRisks can automatically populate information for a breast cancer risk calculation and a new prototype interface shows this data to patients for review and modification. Patients can then add additional personal health information for running updated risk models and aligning each patient’s risk with appropriate preventive interventions. This enhanced version of RealRisks is now the subject of an R21 grant from the National Institute on Minority Health and Health Disparities that aims to conduct user evaluations; assess its effect on patient activation, risk perception, and usability among multi-ethnic high-risk women; and identify multilevel barriers to clinic implementation. The team hopes to empower patients to make deliberate decisions that are appropriate to their risk, make informed decisions, and feel activated in doing so.

Given the modular architecture of RealRisks, the team’s longer-term goal is to continue building it out as a personalized medicine tool that delivers risk-appropriate screening and prevention options depending on the individual user. This has relevance for diseases beyond breast and ovarian cancers. For example, given that many women at high risk for breast cancer are also at high risk for cardiovascular disease, the team has applied for a grant to study the impact of collecting cardiovascular disease risk through RealRisks and how women understand (and juxtapose) their cancer risk and cardiovascular disease risk. In this way, the team continues to enhance RealRisks to be applicable to an expanding number of health conditions that involve preference-sensitive decisions. They also continue to pursue pragmatic studies in real-world clinical practice to facilitate further integration of RealRisks and BNAV into clinic workflow and messaging.